Micro Competition Enzyme Linked Immunosorbant Assay for Human Apolipoprotein B

ANNALS OF CLINICAL AND LABORATORY SCIENCE, Vol. 16, No. 4 Copyright 1986, Institute for Clinical Science, Inc. Micro Competition Enzyme Linked Immunosorbant Assay for Human Apolipoprotein B ARUN U NUNE, P h.d., ROBERT K. NAVIAUX, M.S., JOE C. CHRISTIAN, P h.d., M.D., and DAVID J. GOLDSTEIN,* M.D., Ph.D. Department o f Medical Genetics, Indiana University School o f Medicine, Indianapolis, IN 46223 ABSTRACT A solid phase micro com petition enzyme linked immunosorbant assay (microcelia) was developed to m easure apolipoprotein B (Apo B) in hum an plasma. The soluble Apo B com peted with the solid phase antigen for antibody binding. After washing, alkaline phosphatase labeled goat anti-rabbit IgG antibody was added and the plates w ere w ashed and assayed. The minimal quantifiable concentration of Apo B was one mg per L. This enzym e immunoassay (ElA) yielded values which correlated with values of samples assayed in reference laboratories by radioimmunoassay (RIA) (r = 0.97) and by electroimmunoassay (r = 0.92). The electroim munoassay overestimates the activity of hyperlipem ic samples com pared to m icrocelia. The assay offers advantages over other existing tech niques including short incubation time, high sensitivity, technical simplicity, elimination of radioisotopes, low cost, and use of a universal enzyme label. ease.13 Several of these studies further indicated that the apolipoproteins are b e tte r predictors of heart disease risk than th e lipoprotein cholesterol fractions.13 Fruchart et al8 observed that the ratio of apolipoprotein A -l (Apo A-l) to ap o lip o p ro tein B (Apo-B), u n d e r th e range of 0.98 to 1.38, predicted coronary he^.rt disease with remarkable accuracy. M ethods now used for apolipoprotein quantitation include column chrom atography with subsequent protein determ i nation, acrylam ide gel electrophoresis coupled with quantitative densitometry, an d im m u n o a s s a y s. T h e f ir s t tw o 2 7 8 0091-7370/86/0700-0278 $01.80 Institute for Clinical Science, Inc. Introduction A bnorm al concentrations of plasm a lipoproteins are a major risk factor in the d ev elo p m en t of p re m a tu re ischem ic heart disease.17 C oncentrations of the major lipoprotein fractions have generally been evaluated in terms of lipoprotein cholesterol; recently, 18 epidem iologic studies confirmed the association of apolipoproteins with ischemic heart dis * Address reprint requests to David J. Goldstein, M.D., P h.d., D epartm ent of Medical Genetics, Indiana University School of Medicine, 702 Barnhill Drive, RR129, Indianapolis, IN 46223.

methods require prior separation of lipoproteins from other serum proteins by ultracentrifugation and, hence, are not suitable for use in clinical laboratories. In c o n tra s t, im m u n o a ssa y s can b e applied to whole plasma. Immunoassays include radioim m unoassay, rocket electrophoresis,6,8,16 radial immunodiffusion, im m unonephelom etric and enzym e linked im m unosorbant assay (ELISA).5,7,10,13 The radioimmunoassay requires the use of radioisotopes, a 24 hour or more incubation period, and a gamma counter. Rocket electrophoresis (electroimmunoassay) and radial im m u nodiffusion re q u ire re la tiv e ly large amounts of antibodies and two to three days for completion. The im m unonephelom etric assay depends upon particle size and can be influenced by the presence or absence of other proteins and the size of the immune complex. The micro com petition ELISA (micro- CELIA) for Apo B described here does not require a gamma counter, radioisotopes, or as much time as the radioimmunoassay. It is more sensitive than the rocket electroimmunoassay and requires much less antibody. It is our belief that this assay is a significant step towards a reproducible, inexpensive, convenient assay su itab le for use in clinical and research laboratories. M aterials and Methods P l a sm a S a m p l e s f o r A n a ly sis a n d P u r if ic a t io n PI asm a was o b ta in e d from fa ste d d o n o rs u n d e r sta n d a rd c o n d itio n s.2 Blood was collected in disodium ethylene diam ine tetraacetic acid (EDTA) (one g per L) and prom ptly separated by low speed centrifugation at 4 C. The sam ples w ere stored at 4 C for one to fo u r w eeks or at 20 C for lo n g e r periods. Q uality control sam ples w ere stored at 20 C in sealed glass vials M IC R O C E L IA F O R A P O L IP O P R O T E IN B 2 7 9 until the day of immunochemical analysis. A n t is e r u m P r e p a r a t io n The antibody used was a gift.* This antibody was produced by the following method: purified low density lipoprotein (density range 1.030 to 1.050, L D L 2), prepared by sequential ultracentrifugation9 was emulsified with an equal volume of Freund com plete adjuvant. The rabbits w ere im m unized with four mg of LD L protein. The first two injections, given two days apart, contained com plete Freund adjuvant. A booster dose of two mg protein with incom plete adjuvant was given at one m onth intervals. Rabbits w ere bled on the 7th and 14th days following th e booster injections. The anti-l D L sera used here was isolated from a single rabbit. This antibody has been dem onstrated to be specific for Apo B.9 A p o l i p o p r o t e in B P r e p a r a t io n (LDL, D e n s it y 1.030 t o 1.050 k g p e r L) Narrow density L D L 2 was produced by s e q u e n tia l u ltr a c e n trifu g a tio n.9 Plasma from fasted normolipemic donors was adjusted to a density of 1.030 kg per L with solid KBr and 12 m L fractions were ultracentrifugedf at 40,000 rpm at 5 C for 22 hours. The upper three ml w ere d isc a rd e d. T h e in fra n a te was layered w ith 1.030 kg p er L KBr and recentrifuged. The top th ree ml w ere again discarded. The rem aining fractions w e re p o o le d a n d th e d e n s ity w as adjusted to 1.0050 kg per L with solid KBr. Aliquots w ere layered with a KBr solution (density = 1.050 kg per L) and centrifuged at 40,000 rpm for 24 hr at 5 C. The top three m L w ere saved, aliquoted, layered w ith KBr (density = * Received from Dr. John Albers of die Northwest Lipid Research Laboratory, Seattle, WA. t 50 Ti Beckman Rotor, Palo Alto, CA.

2 8 0 U N U N E, N A V IA U X, C H R IS T IA N, A N D G O L D S T E IN 1.050 kg p er L) and recentrifuged. The upper three ml w ere rem oved as LD L2 and were dialysed for 24 hr against 0.001 m ol p e r lite r sodium ED TA at ph 7. Polyacrylam ide gel electrophoresis of this LD L2 in great excess dem onstrated th e absen ce of alb u m en, apo E and apo C.11 This L D L 2 was assayed for protein by th e m eth o d of L ow ry14 using bovine serum album en (BSA) as a standard. The values o b ta in e d w e re m u ltip lie d by 0.865 (chromogenicity index) to correct for the m easured difference betw een BSA and LDL2.11,15 Sam ples of 0.1 m L w ere sto red at 20 C until used for immunochemical analysis. On the day of assay, the sample was diluted with 0.01 mol per liter phosphate buffered saline (PBS) ph 7.4 (0.2 g KH2P 0 4, 2.17 g N ah P 0 4 heptahydrate, 8.0 g NaCl, 0.2 g KC1, 0.2 g Na Azide p e r L) to give p ro tein concentrations ranging from 1.0 mg p er L to 160 mg per L. P r e p a r a t io n o f V e r y L o w D e n s it y L ip o p r o t e in s ( V L D L ) Fifty Ti tubes w ere filled with plasma from a nonfasted normolipidemic donor, layered w ith 0.15 mol per liter NaCl, and centrifuged^ for 18 hours at 40,000 rpm at 5 C. T he u p p er 2.5 m L w ere rem oved, pooled, and recen trifu g ed after layering w ith 0.15 mol p e r liter NaCl. The top 2.5 m L w ere rem oved and extensively dialyzed against 0.15 mol per liter NaCl, 0.001 mol per liter EDTA, and w ere used as VLDL in competition assays in Apo B coated microtiter plates. W h o l e LD L P r e p a r a t io n Low density lipoprotein (density 1.019 to 1.063 kg p er L) was prepared by sequential ultracentrifugation. The density of the plasm a from a fasted donor was adjusted to 1.019 kg p er L with solid KBr and spun at 40,000 rpm for 20 hr. After centrifugation, the top th ree ml w e re a s p ira te d an d d isc a rd e d. T he infranate was layered with a 1.019 kg per L KBr solution and centrifuged. After centrifugation the top 3 m L were aspira te d and d isc a rd e d. Solid K Br was added to the rem aining fraction to a density of 1.063 kg p er L and recentrifuged for 24 hr. The three ml were aspirated as LDL, 30 fxl of ACG* w ere added, and it was dialyzed against 0.001 mol per liter EDTA ph 7.0 for 24 hr. The LD L was u sed for c o m p e titio n again st Apo B bound microtiter plates. P u r if ic a t io n o f A p o l ip o p r o t e in A - l High density lipoproteins (HDL) of density 1.063 to 1.210 kg p er L w ere isolated from fasted plasm a by sequential ultracentrifugation. G uanidine hydrochloride was added to a six mol per liter concentration and the sample was incubated at 37 C for three h r.15 The sample was dialyzed against one mmol p er liter disodium ED TA at ph 7.4 and then the density was adjusted to 1.210 kg per L with solid KBr. The sample was then centrifuged at 100,000 x g for 26 h r at 5 C. The bottom fraction contains mainly Apo A -l.2 This Apo-1 fraction is dialyzed against affinity buffer (10 mmol p er liter Tris, one mmol p er liter EDTA, 150 mmol per liter NaCl, ph 7.4) and then delipidated with a 1:3 ether:ethanol solution. The resulting delip id ated prep aratio n was added to a DEAE sepharose C1-6B column, equilibrated with 0.03 mol per liter Tris HC1 ph 8.0 containing six mol per liter urea and then eluted with a linear NaCl gradient (0 to 0.125 mol per liter NaCl) to yield purified Apo A -l.2 t L5 50 Beckman Ultracentrifuge Palo Alto, CA. * Sodium azide, 20 g per L; chloramphenicol, 5 g per L; and gentamicin, 10 g per L.

The purity of Apo A -l was confirmed by acrylamide gel electrophoresis according to the m ethod of Davis3 using a 7.5 percent acrylam ide concentration containing eight mol per liter urea. P r e p a r a t io n o f H D L 3 ( D e n s it y = 1.125 t o 1.21 k g p e r L) The fraction rem aining after removal of LD L (density = 1.063 kg p er L) was pooled, th e d en sity was ad ju ste d to 1.125 kg per L by adding solid KBr, and the tubes w ere layered with KBr solution d ensity = 1.125 kg p e r L. T he re su ltin g solution was cen trifu g ed at 40,000 rpm for 26 h r in a 50 Ti Beckman Rotor. The top 3.5 ml w ere discarded. T he infranate was ad justed to density 1.210 kg p e r L w ith solid KBr, was layered with 1.21 kg per L KBr solution, and was recentrifuged. The top 2.5 ml of each tube w ere collected and recentrifuged after layering with 1.21 kg p er L KBr solution. This was repeated and the top 2.5 m L w ere H D L 3. This was dialyzed extensively against 0.15 mol per liter NaCl containing 0.01 mol per liter ED TA.4 This was used for competition for antibody binding in Apo B bound m icrotiter plates. P r e p a r a t io n o f C h y l o m ic r o n s Tubes (SW 50L) w ere two-thirds filled w ith postprandial plasm a and layered w ith 0.15 mol p er lite r of NaCl. The tubes w ere centrifuged at 23,000 rpm for 45 min and the top 0.5 ml was removed. T hese 0.5 m L sam ples w ere pooled, redistributed, layered with 0.15 mol per liter NaCl, and recentrifuged. The top 0.5 m L was the chylom icron fraction, and it was dialyzed extensively against 0.15 mol per liter NaCl containing 0.01 mol per liter EDTA. This was used for M IC R O C E L IA F O R A P O L IP O P R O T E IN B 28 1 competition for antibody binding in Apo B bound microtiter plates. E n z y m e L in k e d I m m u n o s o r b a n t A ssa y: D e t e r m in a t io n o f O p t im a l A p o B C o a t in g C o n c e n t r a t io n a n d A n t ib o d y D il u t io n The wells of m icrotiter plates* w ere coated overnight at 4 C in quadruplicate w ith 200 (xl of various concentrations (0.00, 0.01, 0.03, 0.1, 0.3, 1.0, 3.0, 10, 30, 100 mg per L) of Apo B in 50 mmol per liter carbonate bicarbonate buffer (1.5 2 g N a2c O s p e r L a n d 2.9 3 g NaHCOs p er L), ph 9.6 containing 0.2 g per L sodium azide. After incubation, the coating solution was rem oved and the plates w ere washed three times with 0.01 mol per liter PBS, ph 7.4 containing 0.5 ml p er L Tween 20 and 0.2 g per L sodium azide, and the fluid was aspirated. Coating was also done in PBS and the results w ere com pared. Since there was no significant difference in coating efficiency, the carbonate buffer was used su b se q u e n tly for c o n v e n ie n c e since other antigens used in this laboratory bind more efficiently in carbonate buffer. Coated plates w ere stored for up to two weeks at 4 C w ithout any loss of binding by anti-human Apo B antibody. R abbit anti-hum an Apo B antibody was diluted with 0.01 mol p er liter PBS, ph 7.4 containing 1.0 g per L BSA to produce various dilutions (10 ~2, 10 3, 10~4, 3 X 10-4 ) and 200 xl aliquots were added to respective vertical rows of the coated plates. The plates were incub a te d at 37 C for tw o h o u rs, th e n w ashed as previously stated, and the fluid was aspirated. The second antibody, alkaline phosphatase conjugated goat anti-rabbit IgG f was diluted 1:800 with 0.01 mol per liter PBS ph 7.4 (antibody conjugate). Two hundred fxl aliquots were added to each well, and the plates w ere incubated for * Immulon I, Dvnatech Lai) Inc., Alexandria, VA.

2 8 2 U N U N E, N A V IA U X, C H R IS T IA N, A N D G O L D S T E IN two hours at 37 C, before w ashing as stated previously. Two hundred jll of phosphatase substrate solution (one g per L p-nitrophenylphosphate disodium t in 1.0 mol p er liter diethanolam ine ph 9.8 containing 0.2 g per L sodium azide and 0.1 g p er L MgCl2) w ere added to each well. After sufficient color developm ent, the alkaline phosphatase reaction was stopped by the addition of 50 jll of three mol per liter NaOH to each well in the sequence of addition of substrate. Absorbance was read$ at 405 nm. The mean absorbance was plotted against the log of the coating concentration, the optimal coating concentration (1.0 mg per L), and the optimal antibody dilution for com petition, the dilution that produced 70 percent of the maximal absorbance at the optimal coating concentration, (1:20,000) was determ ined. These concentrations were used in subsequent experiments. D e t e r m in a t io n o f D e t e r g e n t C o n c e n t r a t io n N e e d e d t o M in im iz e N o n s p e c if ic B in d in g produce a higher background. The plate was refrigerated overnight. Next, the vertical rows of each quadrant were washed three times with various concentrations of Tween 20 (0.0, 0.1, 0.3, 0.5, 1.0, or 6.0 ml p er L) in 0.01 mol per liter PBS. Anti-Apo B antibody, diluted 1:20,000 in 0.01 mol per lite r PBS w ith 1.0 g p e r L BSA was added to all except the third quadrant to which only the buffer was added. This quadrant served as a control for nonspecific binding of the antibody conjugate to the Apo B. Next, antibody conjugate was added and, after incubation and washing using various Tween 20 detergent concentrations as m entioned previously, the plate was assayed as described in the previous experim ents. The mean absorbance was plotted versus the various Tween 20 concentrations in the wash solution. (Nonspecific binding was least above 0.3 ml per L Tween 20; 0.5 ml p er L was used in subsequent experim ents for convenience since this concentration was optimal in many other assays.) M icro titer p lates w ere d em arcated into four quadrants. In the first quadrant, 200 xl of carbonate buffer without apolipoprotein was added. This quadrant served as an uncoated control for nonspecific binding. In the second and third quadrants, 200 jll of carbonate buffer containing 1.0 mg per L of Apo B was added. The fourth quadrant was coated with one mg per L Apo A -l and served as a non specific binding, antigen coated control which dem onstrated the absence of Apo A -l immunologic cross reactivity. This control was added because steric hindrance of nonspecific binding might alter the am ount of total antibody bound. If this occurred, an uncoated well would + Sigma Chemical, St. Louis, MO. + Dynatech Micro ELISA Reader. D e t e r m in a t io n o f D e t e r g e n t C o n c e n t r a t io n in t h e F ir s t A n t ib o d y N e e d e d t o O b t a in M a x im a l A n t ig e n A n t ib o d y B in d in g Another plate, coated with one mg per L of Apo B, was incubated with solutions of first antibody (1:20,000 in 0.01 mol per liter PBS with 1.0 g p er L BSA) containing various Tween 20 concentrations (0.0, 0.5, 1.0, 2.0, 3.0, 4.0, 6.0, 8.0, 10, 20, 30, 50, or 100 ml p er L), washed with 0.5 ml per L Tween 20 in 0.01 mol per liter PBS, incubated with antibody conjugate and assayed as described previously. The mean absorbance was plotted versus the various Tween 20 concentra tio n s in th e firs t an tib o d y. (The antigen-antibody reaction was maximal without detergent.)

P r e p a r a t io n o f S t a n d a r d C u r v e f o r A p o B a n d C o m p a r is o n w it h R e f e r e n c e L a b o r a t o r ie s The procedure is outlined in figure 1. M icrotiter plates w ere coated with one mg p er L of Apo B and then washed with 0. 5 m L p er L Tween 2 0 in PBS. One hundred jll of various Apo B concentrations up to 1 0 0 mg per L diluted in PBS w ere added to vertical rows of the plates. N ext, 1 0 0 jil of anti-a po B antibody d ilu te d to 1 :1 0,0 0 0 (fin al d ilu tio n 1 : 2 0, 0 0 0 ) w ere added to each well. The plate was incubated at 37 C for two hr and then w ashed as previously stated. The plate was em ptied, filled with 2 0 0 jl1 of 1 : 8 0 0 diluted antibody conjugate per well, incubated, washed, em ptied, and assayed, as stated previously. Mean absorbance was plotted versus the log of the competing antigen concentration. The absorbance in the absence of added Apo B was considered to be 100 percent. Using this, the data from five experim en ts done on five days w ere compared and plotted as percent inhibition versus the log of the com peting Apo B concentration. M IC R O C E L IA F O R A P O L IP O P R O T E IN B 2 8 3 O th er antigens (chylom icrons, Apo A-l, lipoprotein free media, LDL, and VLDL) in various concentrations (0.1 to 300 ng per L) w ere com pared to Apo B in th e ir ability to com pete for Apo B antibody binding (figure 2). Analytical recovery was done by adding known quantities of Apo B to specific dilutions of plasma and assaying the samples before and after the additions. C o m p e t it io n E n z y m e L in k e d I m m u n o s o r b a n t A ssay o f P l a sm a The wells of m icro titer plates w ere incubated at 4 C overnight with 200 xl of 1.0 mg p er L of Apo B. The plates w ere then washed th ree tim es with 0.01 mol per liter PBS ph 7.4 containing 0.5 ml per liter Tween 20 and 0.2 g per L sodium azide, and the wells w ere aspirated. One hundred xl of various Apo B solutions (0.00, 0.1, 0.3, 1.0, 3.0, or 10 mg per L) w ere added to wells in quadruplicate to generate a standard curve. Secondary standards and plasma samples w ere diluted 1:200 w ith 0.01 mol per liter PBS, and 100 (xl of the diluted FIGURE 1. Schematic representation of the steps involved in a competition ELISA. Ag = antigen (Apo B); AB = rabbit anti-human Apo B antibody; Conjugate = alkaline phosphatase conjugated goat anti-rabbit antibody; Apo B = apolipoprotein B.

2 8 4 U N U N E, N A V IA U X, C H R IS T IA N, A N D G O L D S T E IN COMPETING ANTIGEN CONCENTRATION (m g/l ) F igure 2. The effect of increasing concentrations of various antigens on binding to Apo B coated microtiter plates. samples w ere added to the rem aining wells as quadruplicates. The anti-hum an Apo B antibody was diluted 1:10,000 in PBS and 100 jll w ere added to each well. The plates w ere incubated at 37 C for two hr, washed and em ptied as stated previously. T he goat a n ti-ra b b it IgG antibody conjugated to alkaline phosphatase was diluted 1:800 with PBS, and 200 xl w ere added to each well. The plate was in cu b ated at 37 C, w ashed, and emptied as stated. The alkaline phosphatase bound to the plate was assayed by adding 200 xl of substrate (one g per L p -nitrophenyl phosphate disodium in one mol per liter diethanolam ine, ph 9.8 with 0.2 g per L sodium azide and 0.1 g p er L MgCl2) to each well. After sufficient color developm ent, the reaction was stopped by adding 50 (jll of th ree mol p er liter of N ao H to each well. The absorbance was read at 405 nm. The percent of inhibition was calculated using w ells w ith o u t com peting antigen as 100 p ercent absorbance so that the concentration of Apo B could be d e te rm in e d by com parison w ith th e stan d ard curve. T he re su ltin g values were m ultiplied by a 0.865 chromogenic factor to convert the values obtained by comparison with BSA to Apo B mass.11 15 The concentrations of Apo B in the secondary standards w ere determ ined in four reference laboratories: Dr. Schoenfeld at St. Louis, MO (RIA); Dr. Albers at Seattle, WA (RIA); Sebia at Issy-Les- Moulineaux, France (electroim m unoassay, EIA); and Dr. Stein at Cincinnati, OH (EIA). The values obtained by the m icrocelia described here w ere compared to the values of the reference laboratories by correlation, as has been reco m m e n d e d by th e S ta n d a rd iz a tio n Com m ittee of the International Union of Immunological Societies.18 Results and Discussion Apolipoprotein B adheres to polystyrene m icrotiter plates perm itting assay by microcelia. W hen various concentrations of Apo B in e ith e r carbonate buffer or PBS w ere in cu b ated in the m icrotiter wells, near maximal binding occurred at one mg per L (figure 3). The optim al antibody dilution, the concentration that gave 70 percent of maximal antibody binding, was 1:20,000. Non- APO B COATING CONCENTRATION (fig/ml) F igure 3. The efficiency of binding of various rabbit anti Apo B antibody dilutions (O = 10 2, = 10-3, = 1 0 '4, = 3 X 10 4) to varying coating concentrations of Apo B plotted on a log scale.

specific binding of antibody against Apo B and the antibody conjugate was eliminated with 0.3 ml per L of Tween 20 added to the PBS wash solution (figure 4). F o r co n v e n ie n c e, 0.5 m L p e r L Tween 20 was used for washes in this assay. W hen the effect of various Tween 20 concentrations present during the antibody incubation was examined to determine w hether this detergent inhibited or enhanced antibody binding to Apo B, binding was found to be inversely proportionate to d e terg en t concentration (figure 5). Thus, detergent was not used during this step. Nonspecific binding of antibody conjugate can be quantified w hen the first antibody is absent. In this case, the nonspecific b in d in g of first a n tib o d y to another apolipoprotein (Apo A-l) and the binding of antibody conjugate to Apo B (in the absence of first antibody) gave sim ilar backgrounds (data not shown), M IC R O C E L IA F O R A P O L IP O P R O T E IN B 2 8 5 PERCENT TWEEN 20 IN 1st ANTIBODY FIG U R E 5. The effect of various Tween 20 concentrations on the binding of anti-apo B antibody to microtiter wells coated with Apo B. indicating that the major contribution to background binding is probably due to the nonspecific binding of the antibody conjugate (figure 4). W hen anti Apo B antibody is added to an Apo B-coated well that contains free Apo B, the free Apo B com petes with the bound Apo B for th e added antibody. Less antibody binds to th e well w hen there is more free antigen. This competition (figure 6) is dependent on concentration and perm its quantitation of Apo B in the solution. Apo B concentrations DETERGENT CONCENTRATION (percent) F i g u r e 4. The effect of increasing Tween 20 concentration in the wash solution on antibody binding, the detergent concentration is plotted as log versus mean absorbance for Apo A -l coated and Apo B coated microtiter wells. APO B Concentration (mg/l) F i g u r e 6. Standard calibration curve for Apo B. The effect of increasing Apo B concentration (log scale) Apo B antibody binding to Apo B coated microtiter plates.

2 8 6 U N U N E, N A V IA U X, C H R IS T IA N, A N D G O L D S T E IN above one mg per L can be quantified. The optim al range for q u antitation is betw een th ree and 30 mg p er L. The coefficient of variation among quadruplicate m easurem ents of plasm a samples assayed on the same day was six percent and among mean values m easured day to day was 12 percent. In table I are shown the results of a study of analytical recovery. There is no in te rfe re n c e in th e m e a su re m e n t of know n q u an tities of Apo B ad d ed to plasm a. T he m ean ( + S.D.) p e rc e n t recovery was 101 ± 5 percent. Various lipoproteins contain Apo B, and they com pete with bound Apo B for anti-a po B antibody. C om petition for anti-apo B antibody by added lipoprotein free medium, chylomicrons, Apo B, VLDL, Apo A -l, and LD L is shown in figure 1. The lack of com petition of Apo A -l and lipoprotein free m edia (LPFM) indicates the absence of cross reactive material in these samples. The presence of Apo B in chylomicrons, VLD L and LD L is dem onstrated by their ability to com pete for anti-apo B antibody. The proportion of Apo B in the VLDL and chylom icron fractio n s a p p e a rs to be about one tenth that of LD L and Apo B in this assay. Plasm a D i l u t i o n T A B L E I Analytic Recovery Using Apo B MicroCELIA A dded Apo B (m g /L ) n M e a su re d A po B (m g /L ) R e c o v e r y ( P e r c e n t) 10-4 0.0 8 0.22 0.3 2 0.56 113 1.0 2 1.18 96 3.0 2 3.10 96 10-5 0.0 8 0.0 0.3 2 0.3 100 3.0 2 3.2 107 10 7 0.0 4 0.0 6.0 4 6.0 100 10.0 4 10.0 100 12.7 4 13.0 102 16.7 4 16.5 99 M etho d T A B L E II Comparison of Means from Reference Laboratories with Apo B MicroCELIA n Mean R e fe r e n c e L a b o r a to r y ± SD CELIA C o r r e l a t i o n s r S lo p e RIA samples 4 99 ± 31 94 ± 30 0.97 0.96 EIA samples Secondary 4 129 ± 57 148 ± 65 0.92 0.81 Routine 26 173 ± 43 161 ± 51 0.75 0.77 Hyperlipemic 6 291 ± 93 182 ± 60 0.88 1.38 O u r resu lts (m ean d eterm in atio n s) were com pared in table II and figure 7 with the results of four reference laboratories (four samples done by RIA and 36 by E l A) and correlations w ere calculated. The correlations are summarized in table II. The correlation of the RIA w ith th e C E L IA in d icates th a t th ese methods are nearly identical for estimation of Apo B in plasm a sam ples. The correlation of the E l A with the CELIA in d ic a te s th a t th e y a re c o m p a ra b le except when estim ating Apo B in turbid (hyperlipemic) samples which are consistently higher as m easured by EIA. The system atic overestim ation of Apo B by EIA as indicated by the correlation and slope of the hyperlipem ic sam ples has b e e n n o te d by u sin g n e p h e lo m e tric immunoassay versus CELIA and RIA.18 The ro u tin e sam ples (figure 7) w ere assayed once by the reference laboratory and may indicate the scatter obtained in routine EIA. The mean values and standard deviations, d e te rm in e d by this microcelia, w ere 1.16 ± 0.34 g per L for 31 females and 1.28 ± 0.44 g per L for 32 males. These results are comparable to those means (range 0.81 to 1.59 g per L) reported by Rosseneu et al.19 Although there are num erous available immunoassays for determ ination of Apo B, this assay has some advantages. Radioimmunoassay requires the use of radioisotopes, long incubation periods, and a gamma counter. The electroim -

M IC R O C E L IA F O R A P O L IP O P R O T E IN B 2 8 7 o u / A / a? lb * I 100 8» / E L E C T R O - IMMUNO ASSAY R A D IO - IMMUNO ASSAY 7 200 % D S e c o n d a ry samples R o utine samples O Hyperlipem ic samples A S e c o n d a ry samples I 300 T 400 F i g u r e 7. C om parison of CELIA values with those of various reference laboratories. Correlation coefficients (r) with RIA (0.97), E IA se c o n d a ry samples (0.92), EIA routine samples (0.75), EIA H y p erlip em ic sam ples (0.88). Dashed line represents theoretical slope of correlation. REFERENCE LABORATO RY m unoassay and radial immunodiffusion re q u ire two to th re e days and large am ounts of antibody (ten tim es m ore than RIA). The im m unonephelom etric assay requires relatively large quantities of antibody and is sensitive to the presence or absence of other proteins and the size of the im m une complex. Fruchart and o th ers5,7,13 described oth er en zy m e-lin k ed im m unoassays. O n e 13 w as a c o m p e titio n e n z y m e -lin k e d im m u n o so rb an t assay (CELIA ) using five-fold larger volumes and longer incubation times than in this report. The second5 was a noncom petition ELISA using im m u n o g lo b u lin -b o u n d p o ly sty re n e balls. This sandwich assay has a ten-fold greater sensitivity than the competition assay, but it requires an extra assay step or the purchase of commercially p re pared balls and antibodies. The third7 is a sandwich assay using flexible m icrotiter plates. Antibody from two different animal species is required, b u t it is a very precise and sensitive assay. A fourth enzy m e im m u n o assay, r e p o rte d by H olm quist,10 requires cellulose nitrate coating and conjugation of enzym e to LD L particles; however, it has a ten-fold increased sensitivity com pared with the assay described here. T hree of these immunoassays5,10,13 require more antibody, reagents, and tim e than the assay described here. In addition, sensitivity is not a problem since plasma needs to be d ilu te d 1:200 for th is m icro C E L IA. A nother CELIA using coated beads and tubes19 is similar to this assay. Because it has some advantages over many of these assays, the microcelia is a useful alternative assay for quantifying Apo B. A sim ilar assay for Apo A -l is now being developed in our laboratory, and it is anticipated that a combination of these assays will provide a satisfactory m ethod for m ass sc re e n in g of in d iv id u als to determ ine cardiovascular risk and for the study of the effects of various therapeutic interventions. Acknowledgments Thanks are extended to Pao-Lo Yu, Ph.D. and Lynne Jeffers-Hallock for their assistance with the statistical analysis, and to Jeff Goldstein and Cate Antley for their technical help. This work was supported by the Indiana Affiliate of the American Heart Association, PHS Training Grant G-T-32-DE-07043, PHS HL29090 and the J. W. Riley Foundation. References 1. A l b e r s, J. J., V e n e r a c i a n, G. C., and H a z - ZARD, W. R.: Immunoassay of hum an plasma apolipoprotein B. Metabolism 24:1339-1351, 1975, 2. A l b e r s, J. J., W a h l, R. W., C a b a n a, V. G., H a z z a r d, W., and H o o v e r, J. J.: Quantitation

2 8 8 U N U N E, N A V IA U X, C H R IS T IA N, A N D G O L D S T E IN of apolipoprotein A -l of hum an plasma high density of lipoprotein. Metabolism 25:633-644, 1976. 3. D a v is, B. J.: Disc electrophoresis. II. Method and application to human serum proteins. Ann. N.Y. Acad. Sci. 121:404-427, 1964. 4. D e L a l l a, O. and G O FM A N, J. W.: Ultracentrifugal analysis of serum lipoproteins. Methods of Biochemical Analysis, vol. 1. Delson, D., ed., New York, Interscience, 1954, pp. 459-478. 5. F r u c h a r t, J. C., D e s r e u m a u x, C., D e w a i l l y, P., S e z i l l e, G., J a i l l a r d, J., C a r l i e r, Y., B o u t, D., and C a p r o n, A.: Enzyme immunoassay for human apolipoprotein B the major protein moiety in low-density and very-low-density lipoproteins. Clin. Chem. 24:455 459, 1978. 6. F r u c h a r t, J. C., D u t h i l l e u l, P., K o r a, I., D u q u e s n e, M., C l a v e y, V., and G e n t i l i n i, J. L.: Rapid determ ination of hum an plasma apolipoprotein B. Clin. Chim. Acta 114:123-126, 1981. 7. F r u c h a r t, J. C., F i e v e t, C., O u v r y, D., K o f - FIGAN, M., BEU CLER, I., AYRAULT-jARRIER, M., B a r r y, M. D., and M a r c o v i n, S.: Enzvm e- linked immunoassay on m icrotitre plates for human apolipoprotein B, La Ricerca. Clin. Lab. 14:569-574, 1984. 8. F r u c h a r t, J. C., K o r a, I., C a c h e r a, C., C l a v e y, V., D u t h i l l e u l, P., and M o s c h e t t o, Y.: Simultaneous measurement of plasma apolipoprotein A-l and B by electroimmunoassay. Clin. Chem. 28:59 62, 1982. 9. H a v e l, R. J., E d e r, H. A., and B r a g d o n, J. H.: The distribution and chemical composition of ultracentrifugally separated lipoproteins in human serum. J. Clin. Invest. 34:1345 1353, 1955. 10. H O LM Q U IST, L.: Quantitation of human serum apolipoprotein B bv enzvme immunoassav. Clin. Chim. Acta 12:327-336, 1982. 11. L e e, D. M. and A l a u p o v i c, P.: Composition and concentration of apolipoproteins in very low and low density lipoproteins of normal human plasma. Atherosclerosis i0:501-520, 1974. 12. Lipid Research Clinics Program Manual Laboratory Operations, D H EW Publication (NIH) 1, 75-628, 1974. 13. L i p p e l, K. : Proceedings of the Workshop on A polipoprotein Q uantification, S ep tem ber 2 1-2 2, 1982. (D H H S-N IH publication No. 83-1266, 1983). 14. L o w r y, O. H., R o s e b r o u g h, N. J., F a r r, A. L., and RANDALL, R. J.: Protein measurem ent with the Folin phenol reagent. J. Biol. Chem. 193:265-275, 1951. 15. N a k a i, T., T a m a i, T., Ya m a d a, S., K o b a y a s h i, T., H u y a s h i, T., K u t s u m, Y., O i d a, K., and T a k e d a R.: Plasma lipids and liproproteins of Japanese adults and umbilical cord blood. Artery 9:132-150, 1981. 16. R e a r d o n, M. R., P o a p s i, M. E., U f f e l m a n, K. D., and S t e i n e r, G.: Improved method for quantification of B apolipoprotein in plasma lipopro teins bv electroim m unoassay. Clin. Chem. 27:892-895, 1981. 17. R h o a d s, G. C., C u l b r a n d s e n, C. L., and K o g a n, A.: Serum lipoproteins and coronary heart disease in a population study of Hawaii Japanese men. New Engl. J. Med. 294:293-298, 1976. 18. R o s s e n e u, M., V e r c a e m s t, R., S t e i n b e r g, K. K., and C o o p e r G. R.: Some considerations of methodology and standardization of apolipoprotein B immunoassays. Clin. Chem. 29:427 433, 1983. 19. V a n d e r H e i d e n, G. L., S l a s s e, E. A., Y o r d e, D. E., M a d i e d o, G., and B a r b o r i a k, J. J.: Examination of a com petitive enzvme-linked immunoassay (CELIA) technique and a laser nephelometric immunoassay technique for the measurement of apolipoprotein B. Clin. Chim. Acta i 35:209-218, 1983.